Comment by credit_guy | Hacker News Reader

Comment by credit_guy | Hacker News Reader

Parent Full thread

It so happens that NuScale went public exactly 6 months ago. Since then S&P500 went down more than 15% while NuScale went up more than 30%.

As for Boom, I have to admit, they have a problem.

The problem is that companies like GE that make jet engines, can make quite some nice profits by selling to the military. The XA100 [1] most likely exceeds any requirements Boom can only dream of, but being military technology, it is off limits for them. If GE lands a fat contract to provide engines for the F35, then they'll sell thousands of engines, with very little uncertainty. Getting involved with Boom will only complicate their life. If they don't fully compartmentalize the two departments, they can get in trouble. If they do, there's no synergy. Plus, the doubts will always be there that they leak military technology to the civilian side.

That's a shame. GE makes some incredible jet engines.

Maybe the war in Ukraine will end and one day Putin won't be president anymore, and Russia won't be under sanctions. Then Boom could buy an engine from Saturn, such as [2], the engine that powers various Sukhois. The military-civilian tension will still be there, but Russia will need to export things, so maybe a deal could be made.

[1] https://www.geaviation.com/propulsion/military/xa100

[2] https://en.wikipedia.org/wiki/Saturn_AL-31

A military engine is unlikely to be the answer for Boom. They just won't be cost effective.

The math for efficiency and maintenance is just different for sovereign governments, maybe a derivative could be made more efficient with less maintenance required.

Fighter jets require like 6 hours of maintenance for every hour of flight - sure, it's not just the engines, but it's a big part of it.

The real problem boom has is the number of viable flights is _very_ low - really just the ones Concorde flew (NY-LON, NY-PAR), and Concorde never made money.

A couple comments up someone claimed DXB-NYC being a route - more than half of the great circle route is over land, and it's more than 50% longer than the rosy projections Boom put out. Realistically, they'll be lucky if they're within 10% of their stated range. But of course, not without an engine.

This is what 4,250nm would get you from JFK: http://www.gcmap.com/mapui?R=4250nm%40JFK

We also didn't care about being green back during the Concorde - but Supersonic flights will likely be 5x-10x worse on a per passenger seat mile basis for emissions. Really hard to get around wind resistance being a square of velocity.

> Really hard to get around wind resistance being a square of velocity.

That's a common misunderstanding. Very common.

In cruise mode Concorde was as efficient as a Boeing 747.

When you go faster, you do two things: 1. reduce the angle of attack, 2. fly higher. Concorde was flying twice as high as subsonic jets. The end result is that you still get increased drag, but much less than the square (or cube) of velocity would suggest.

Ok, but increased drag it is, right? Right, but you go faster. If your drag goes up by a factor of 2, but your speed by a factor of 4, then in the end you burn less fuel. It so happens that Concorde's drag went up virtually by the same factor as its speed.

The problem with Concorde was that it was horrible aerodynamics during the takeoff stage. At that point the plane is the heaviest (its has full fuel tanks), so the problem was compounded by some sort of "tyranny of the rocket equation".

Can Boom solve this problem? My guess is that this was their pitch to investors. In the 50+ years since Concorde was designed, both computers and computational fluid dynamics experienced huge advances. The computer I'm writing this on has probably performed more flops since I started writing this comment than the entire computational modeling for Concorde in its decade-long work.

As for military engines (and their maintenance) being expensive. F-35 is like a Formula 1 car, as USAF general Charles Brown once said. Its engines are pushed to the limit on a constant basis. A commercial jet is very different. It does not need to pull huge Gs. Its trajectory is very, very predictable. I'm sure a company like GE could make a very cost efficient jet engine for Boom. It's just that they have much better ROI if they build engines for F-35.

> In cruise mode Concorde was as efficient as a Boeing 747.

Yes - a plane that carries 550 passengers vs a plane carrying 100.

Also, the game has changed a fair bit - no one is building 4 engined planes anymore - because while it was about as efficient as a 747, it’s going to be less than half as efficient as a 777. That carries 400 passengers.

By efficiency I meant the lift-to-drag ratio [1].

The wiki page I linked quotes an L/D ratio for Concorde (in cruise mode) of 7 to 7.5 and for Boeing 747 of 15.3 to 17.7. Taking the mid-points of 7.25 and 16.5 we get that the L/D ratio for Boeing is 2.275 higher than of Concorde. But Concorde's speed was 2.41 times higher than Boeing's (Mach 2.05 vs 0.85).

There are 2 things to go from here to actual fuel per passenger*mile. How much fuel do you need to produce the same thrust, and how much aircraft do you need to carry a passenger. On both counts, supersonic airplanes have a problem: supersonic engines have a lower bypass ratio, so they are inherently less efficient, and supersonic planes need to be sturdier, and therefore heavier than subsonic ones.

These problems are hard.

But the difficulty of building a supersonic aircraft does not come from the air resistance being the square of the speed.

By the way, the L/D ratio decreases with the speed, but it does not tend to zero in the limit. This fact is actually stunning, if you think of it. As the speed goes up, at some point the L/D ratio stops decreasing, so you end up burning significantly less fuel per mile. That's the reason some militarizes invest in hypersonic missiles: they have much longer range than their size alone would make you think.

[1] https://en.wikipedia.org/wiki/Lift-to-drag_ratio

Lift to drag is interesting - but without dividing by seats or tons is incomplete. You’re not really measuring work being done (actually transporting goods or people).

These problems are hard.

Lift induced drag goes down as velocity decreases, parasitic drag goes up as velocity increases. There’s there’s wave drag to contend with. Yes there’s weird places where going faster is helpful, but I doubt we’re going to see real net efficiency gains going faster than Mach .8-9 anytime soon.

Glad we got to this point.

So, can Boom compete with subsonic airplanes. Where can they have an edge?

They have two actually:

1. they are a startup. Boeing and Airbus are old and ossified. Smaller manufacturers are probably not as ossified, but they already have a niche, so their risk appetite is different: they prefer to put their research money to work towards incrementally improving their current designs. Boom is in a life and death situation, they have a huge fire under their feet, and so they'll get more research done for the same amount of dollars

2. Clean slate. All current civilian airplanes are cylinders with flat wings. It's a proven formula. Manufacturers don't want to move too much from a proven formula. But Boom does not have this type of fetters. In particular, they can design a body that follow the Whitcomb area rule [1]. Which is what they are actually doing [2]:

  Contoured fuselage: According to the principle of area-ruling, Overture’s fuselage has a larger diameter toward the front of the aircraft and a smaller diameter toward the rear. Boom has applied this design technique to minimize drag and maximize fuel efficiency at supersonic speeds.

In addition to that, Boom plans to use wings that are not flat (they call them gullwings). Presumably, they know what they are doing, they say they used 26 million-core hours of compute between the first and the second iteration of their Ouverture.

Will the manufacturing costs explode? I guess they should be higher than for traditional geometries. But that works to their advantage: it keeps the current manufacturers from trying these non-standard geometries, and keeps them in a local optimum from which they find it harder to move. But it won't be a dealbreaker for a premium aircraft. That's basically the Tesla Roadster gameplan.

[1] https://en.wikipedia.org/wiki/Area_rule

[2] https://boomsupersonic.com/news/post/boom-supersonic-reveals...

Reply on news.ycombinator.com